Gentoo Forums
Gentoo Forums
Gentoo Forums
Quick Search: in
Gentoo strategic direction re: udev
View unanswered posts
View posts from last 24 hours

Goto page Previous  1, 2, 3, 4  Next  
Reply to topic    Gentoo Forums Forum Index Gentoo Chat
View previous topic :: View next topic  
Author Message
Bones McCracker
Veteran
Veteran


Joined: 14 Mar 2006
Posts: 1564
Location: U.S.A.

PostPosted: Sun Feb 10, 2013 8:38 pm    Post subject: Reply with quote

I think it's a simple fact that the systemd / udev development team cannot be relied upon to permanently keep udev cleanly separable from systemd, because Poettering has clearly stated he wants to eventually drop support for non-systemd employment of udev.

The essence of this conflict, as I see it, is that there are a group of people who think the entire Linux world (less embedded) should standardize on systemd, another group of people who don't want to do that, and the former group is in control of udev and have no intention of forever providing it for use independent of systemd (at least not packaged as a cleanly discrete and independent product).

Poettering may have recently begun to make conciliatory statements to the contrary, and taken some at least temporary actions to assuage the concerns of the second group, but has he ever made a definitive statement to the contrary? If he has, I have not heard it, but I'm not aware of everything that's been said.
_________________
True Liberals are individualists. Democrats, on the other hand, are authoritarian collectivists.
Back to top
View user's profile Send private message
v_andal
Guru
Guru


Joined: 26 Aug 2008
Posts: 302
Location: Germany

PostPosted: Mon Feb 11, 2013 8:23 am    Post subject: Reply with quote

I feel, that the problem is even deeper than this. There's still group of people that want to understand and control their Linux system, and there's another (big) group of people, who want to make Linux alternative to Windows, in other words "user friendly", "simple to use" etc. It is not possible to go both ways at the same time. Sooner or later the split has to happen. The code produced by the advocates of the second group becomes more and more complex and less and less controllable. Their system effectively becomes Lindows, despite being open source. I don't say it is bad. It's just not what I personally want. That's the main reason why I use Gentoo and not Ubuntu. I still have a feeling of control over the system. I don't want user-friendly, I want stuff that I can tailor to my desires :)
Back to top
View user's profile Send private message
ulenrich
Veteran
Veteran


Joined: 10 Oct 2010
Posts: 1136

PostPosted: Mon Feb 11, 2013 9:48 am    Post subject: Reply with quote

Gentoo users should buy weapons to defend their true community ?
Come on,
do listen more to Bill Maher, John Stewart and less Rush Limbaugh and Dennis Prager type of agitations.
Get your feet back on the ground!

There I put the video link:
http://forums.gentoo.org/viewtopic-p-7243360.html#7243360
where Lennart officially denounces he will ever attempt to integrate the Linux kernel into Systemd, because he admits there are use cases of Linux without Systemd and probably ever will be. I must admit beside these two sentences, the rest of the lecture he argues the other way round ...
_________________
fun2gen2
Back to top
View user's profile Send private message
smartass
Apprentice
Apprentice


Joined: 04 Jul 2011
Posts: 189
Location: right behind you ... (you did turn around, didn't you?)

PostPosted: Mon Feb 11, 2013 5:10 pm    Post subject: Reply with quote

What a joke. Poettering (in the video) first says he wants to bring stuff from other unices (FreeBSD), where everything is developed in an integrated fashion. Isn't one of the main Unix mantras "make separate tools that do one job and do it well"?
Back to top
View user's profile Send private message
ulenrich
Veteran
Veteran


Joined: 10 Oct 2010
Posts: 1136

PostPosted: Mon Feb 11, 2013 6:42 pm    Post subject: Reply with quote

journalctl
localectl
loginctl
systemctl
systemd
analyze
ask-password
systemd-cat
systemd-cgls
systemd-cgtop
coredumpctl
systemd-delta
detect-virt
inhibit
id-setup
notify
nspawn
bridge
tmpfiles
password-agent

And for everything a man page. And for every unit type a man page. And for special generator type units. More than 200 pages altogether. These are very long manuals most of it. It is so much help available that systemd-198 will have a table-of-content.man page to get a grip where to find what info.

Nobody of you really looked at systemd before talking (*). Absurdly nearly everything talked about systemd in the Gentoo forum, despite its fame of being the site of experts, is FUD.

@smartass, you missed my twinkling eye joke of the last sentence?

(*)PS: which is possible without cost: systemd and openrc is no problem to install in parallel.
_________________
fun2gen2
Back to top
View user's profile Send private message
Ant P.
Advocate
Advocate


Joined: 18 Apr 2009
Posts: 2280
Location: UK

PostPosted: Mon Feb 11, 2013 7:29 pm    Post subject: Reply with quote

ulenrich wrote:
journalctl
localectl
loginctl
systemctl
systemd
analyze
ask-password
systemd-cat
systemd-cgls
systemd-cgtop
coredumpctl
systemd-delta
detect-virt
inhibit
id-setup
notify
nspawn
bridge
tmpfiles
password-agent

And for everything a man page. And for every unit type a man page. And for special generator type units. These man pages are very long. It is so much help available that systemd-198 will have a table-of-content.man page to get a grip where to find what info.

Nobody of you really looked at systemd before talking (*). Absurdly nearly everything talked about systemd in the Gentoo forum, despite its fame of being the site of experts, is FUD.


You say that as if a single manpage that takes thirty times as long as the corresponding OpenRC equivalent to get to the damn point is a good thing.

If I wanted to spend five hours reading documentation in order to use my system, I'd go install LFS. Their docs would teach me something worth the time.
Back to top
View user's profile Send private message
ulenrich
Veteran
Veteran


Joined: 10 Oct 2010
Posts: 1136

PostPosted: Mon Feb 11, 2013 7:51 pm    Post subject: Reply with quote

Ant P. wrote:
You say that as if a single manpage that takes thirty times as long as the corresponding OpenRC equivalent to get to the damn point is a good thing.

If I wanted to spend five hours reading documentation in order to use my system, I'd go install LFS. Their docs would teach me something worth the time.
To get to the point you need just 6 words:
Just run it, it will work. (This sentence will be criticized as Winass kind of habit, but it is true)

But if you ask for documentation for all the extra features: There are more than 30 multiples of the openrc features. This minimalism can be seen as the true virtue of openrc though!

A few weeks running systemd - but not yet looked into it, it was sufficient for me to know just these:
systemctl status SERVICE
systemctl start/stop SERVICE
systemctl enable/disable SERVICE
systemctl isolate TARGET

One thing I promise you: If you will have a little look into a systemd man page, you will feel kind of enlightened what is possible with Linux.
_________________
fun2gen2


Last edited by ulenrich on Mon Feb 11, 2013 8:03 pm; edited 1 time in total
Back to top
View user's profile Send private message
depontius
Advocate
Advocate


Joined: 05 May 2004
Posts: 2439

PostPosted: Mon Feb 11, 2013 8:02 pm    Post subject: Reply with quote

ulenrich wrote:
Ant P. wrote:
You say that as if a single manpage that takes thirty times as long as the corresponding OpenRC equivalent to get to the damn point is a good thing.

If I wanted to spend five hours reading documentation in order to use my system, I'd go install LFS. Their docs would teach me something worth the time.
To get to the point you need just 6 words:

Just run it, it will work.
(This sentence will be criticized as Winass kind of habit, I know, but it is true)


Absolutely right - 90% of the time. It's that last 10% of the time that makes it an absolute mess and pain. As for reading the docs, as far as I can tell the detail docs are written to create developers, not to help people hack their own systems. (This is my general criticism of anything from freedsktop.org, not just systemd.)

The counter-example is good old xfree86. For all of the criticism of the arcane, horribly long xf86config file, I remember hacking my own in the bad old days. I was able to read docs and figure out enough to do to get my system working. Arcane, yes... Discoverable, yes... Hackable, absolutely - and that's what counts.

The real problem here is that "Just Works (TM)" software that simply does the right thing 90-95% of the time, frequently gets in the way of your own hacking your way to working, that last 5-10% of the time.

I would be perfectly happy with 90-95% Just Works if it had hackable docs for that last 5-10% instead of become-a-developer docs.

ulenrich wrote:

But if you ask for documentation for all the extra features: There are more than 30 multiples of the openrc features. This minimalism can be seen as the true virtue of openrc.

A few weeks running systemd - but not yet looked into it, it was sufficient for me to know just these:
systemctl status SERVICE
systemctl enable/disable SERVICE
systemctl isolate TARGET

_________________
.sigs waste space and bandwidth
Back to top
View user's profile Send private message
ulenrich
Veteran
Veteran


Joined: 10 Oct 2010
Posts: 1136

PostPosted: Mon Feb 11, 2013 8:12 pm    Post subject: Reply with quote

depontius wrote:
ulenrich wrote:
Just run it, it will work.

Absolutely right - 90% of the time. It's that last 10% of the time that makes it an absolute mess and pain. As for reading the docs, as far as I can tell the detail docs are written to create developers, not to help people hack their own systems. (This is my general criticism of anything from freedsktop.org, not just systemd
Then you are more into it than me at this time. Would be nice for me to here more about your problems in this forum, because it is a fine way for me to go further, which I plan with systemd.

Yes, the problem understanding man pages. They help as a repetition when you don't remember but you understood once a time ago. We should write some more Systemd wiki pages? Arch should have quiet well as they do ever ?
_________________
fun2gen2
Back to top
View user's profile Send private message
Bones McCracker
Veteran
Veteran


Joined: 14 Mar 2006
Posts: 1564
Location: U.S.A.

PostPosted: Mon Feb 11, 2013 8:22 pm    Post subject: Reply with quote

I detect a faint waft of Kool Aid.
_________________
True Liberals are individualists. Democrats, on the other hand, are authoritarian collectivists.
Back to top
View user's profile Send private message
ulenrich
Veteran
Veteran


Joined: 10 Oct 2010
Posts: 1136

PostPosted: Mon Feb 11, 2013 8:43 pm    Post subject: Reply with quote

BoneKracker wrote:
I detect a faint waft of Kool Aid.
"faint waft" - sounds like scots gaelic to me. Is it a Craig Ferguson kind of twinkle?
_________________
fun2gen2
Back to top
View user's profile Send private message
Dr.Willy
Guru
Guru


Joined: 15 Jul 2007
Posts: 324
Location: NRW, Germany

PostPosted: Mon Feb 11, 2013 10:00 pm    Post subject: Reply with quote

BoneKracker wrote:
I detect a faint waft of Kool Aid.

ulenrich is the resident systemd evangelist, didn't you know?
Back to top
View user's profile Send private message
krinn
Advocate
Advocate


Joined: 02 May 2003
Posts: 4150

PostPosted: Mon Feb 11, 2013 10:28 pm    Post subject: Reply with quote

ulenrich wrote:

Nobody of you really looked at systemd before talking (*). Absurdly nearly everything talked about systemd in the Gentoo forum, despite its fame of being the site of experts, is FUD.


No, i won't try or install systemd. I don't know if systemd is good or bad, and i don't care at all.

It's about a rampant takeover done by a company using a well know scheme : if you want your product to be the one to get, make it the standard : at end all products will be the same and users will think "why i should use debian to be like redhat ? I will just use redhat"

If redhat want their system, they can just do what they want with their distro, but they will endup with a special distro, with many things too specific that will need a special implementation of program in order to make them run on it : a dead end -> You're the one, but sadly you are alone. (a dead end for opensource, a real good move for closed source like microsoft has done)

It's way better to tiedup linux and its users with chains, slowly going in jail. This way you will have compatible program that doesn't need to be adapt for your distro, all the market at your hands -> Even you aren't alone now (many distros did follow you), you're the one as others are just followers.

And this is what they appears to do : udev, systemd, pulseaudio, kit: tied them all, so everyone MUST use and have them in their distro, and when all distros endup the same, users will knows the one to use.
Won't be that easy for gnome. Hmmm but not undoable... https://mail.gnome.org/archives/desktop-devel-list/2011-May/msg00427.html

So if to replace the "so-called" broken tools that works for years with new shiny kickass programs that run at light speed, is stable and everything you wish to endup with chains, i prefer keep my borked tools and freedom. systemd is not intended to be portable...

I don't say i'm saying truth, as i don't work for redhat and i'm not poettering. But it's the way i see those wanted vertical integration and i don't like it.
Back to top
View user's profile Send private message
duby2291
Guru
Guru


Joined: 17 Oct 2004
Posts: 426

PostPosted: Tue Feb 12, 2013 12:51 am    Post subject: Reply with quote

ulenrich wrote:


And for everything a man page. And for every unit type a man page. And for special generator type units. More than 200 pages altogether. These are very long manuals most of it. It is so much help available that systemd-198 will have a table-of-content.man page to get a grip where to find what info.

Nobody of you really looked at systemd before talking (*). Absurdly nearly everything talked about systemd in the Gentoo forum, despite its fame of being the site of experts, is FUD.


You say this as if your trying to make it sound like a good thing....

200 man pages for something that it primary focus is to start and stop services..... Thats what is absurd.

Quote:
But if you ask for documentation for all the extra features: There are more than 30 multiples of the openrc features. This minimalism can be seen as the true virtue of openrc though!

A few weeks running systemd - but not yet looked into it, it was sufficient for me to know just these:
systemctl status SERVICE
systemctl start/stop SERVICE
systemctl enable/disable SERVICE
systemctl isolate TARGET


Then that is literally all it -should- be. It should start and stop services.... period.... But wait sysvinit already does that.... ooops....
_________________
MB: Biostar TForce 6100 AM2 @ 250x10
CPU: AMD Athlon 64 3800+ X2 @ 2500mhz
MEM: G. Skill DDR2-800 2GB @ DDR2-1000
GPU: nVidia GeForce 7600 GT
OS: Gentoo Linux 2006.1
Back to top
View user's profile Send private message
smartass
Apprentice
Apprentice


Joined: 04 Jul 2011
Posts: 189
Location: right behind you ... (you did turn around, didn't you?)

PostPosted: Tue Feb 12, 2013 7:05 am    Post subject: Reply with quote

ulenrich wrote:
journalctl
localectl
loginctl
systemctl
systemd
analyze
ask-password
systemd-cat
systemd-cgls
systemd-cgtop
coredumpctl
systemd-delta
detect-virt
inhibit
id-setup
notify
nspawn
bridge
tmpfiles
password-agent

ulenrich, if all these tool were able to work independently (e.g. I could replace journald with a cron of my choice), then I'd consider systemd being the "unix way"
Back to top
View user's profile Send private message
ssuominen
Developer
Developer


Joined: 30 Sep 2005
Posts: 2108
Location: Finland

PostPosted: Tue Feb 12, 2013 7:09 am    Post subject: Reply with quote

Why are we talking about systemd in a thread labeled as "udev"? Current technical facts are such that they are not tied together in a way it should not bother anyone. If the maintainers
are happy with it's build system, that covers the only "problem" -- as in, not a problem at all.
Back to top
View user's profile Send private message
Yamakuzure
Veteran
Veteran


Joined: 21 Jun 2006
Posts: 1347
Location: Bardowick, Germany

PostPosted: Tue Feb 12, 2013 11:40 am    Post subject: Reply with quote

ssuominen wrote:
Why are we talking about systemd in a thread labeled as "udev"? Current technical facts are such that they are not tied together in a way it should not bother anyone. If the maintainers
are happy with it's build system, that covers the only "problem" -- as in, not a problem at all.
hear, hear!

I can tell from my own systems, that udev-197-r4 works flawlessly with openrc-0.11.8 and sysvinit-2.88-r3. But: I do not have a separate /usr partition. I do not need one.

As for systemd, ulenrich, I have never tried it. I read into it a bit, but decided, for me, that it is too early to try out. Generally it sounds like a good idea. Unification and simplification of related tasks is always better than separatism of the seemingly same into different parts. But some understand "the unix way" to mean "as complex as possible for the sake of it."
On the other hand, "simplification" does not mate too well with "over 200 man pages". ;) And, of course, like smartass wrote, the lack of choice to substitute provided tools with tools of your liking counts against systemd. At least for now.
_________________
I *do* know that I easily aggravate people due to my condensed writing. Rule of thumb: If I wrote anything that can be understood in two different ways, and one way offends you, then I meant the other! ;)
Back to top
View user's profile Send private message
krinn
Advocate
Advocate


Joined: 02 May 2003
Posts: 4150

PostPosted: Tue Feb 12, 2013 12:47 pm    Post subject: Reply with quote

ssuominen wrote:
Why are we talking about systemd in a thread labeled as "udev"? Current technical facts are such that they are not tied together in a way it should not bother anyone. If the maintainers
are happy with it's build system, that covers the only "problem" -- as in, not a problem at all.


Because it's gentoo chat, about "strategic direction" for udev.
If it were gentoo bugs, we would be off topic speaking about systemd, but handling udev is tied to other subjects like eudev, systemd, redhat, pottering ...
It we speak about having children, it's not off topic to speak about getting a wife :)

And the main focus is there, you've told us udev isn't tied with systemd so we could still use it as-is. That's a short term strategy: closing eyes to what lennart says and wait until...
I think it's not bad if gentoo devs tell us what will happen after that "until" is reach. I see eudev as a more safety bet because at least it cover that case, and it might even be a never finish product but could influence udev (and i do think it has already) future.

I would like to here from you, what you will do then ? Because for now, you appears to be a key in gentoo about udev, and also hostile to eudev. You may not, but appears really. It would be more sane if you were having a plan mid or long term and share it.
Right now, the short term looks dirty, as you can see with the latest udev stabilization and networking renaming troubles everyone get. And i didn't test it, but this change looks strange : why would naming my network interface by bus place is better than mac ? There's more chance i will move my ethernet card from on slot to another than i would add another card with the same mac. High havock <> benefits ?
You may didn't notice it, but many users are trying to get off udev by mdev, eudev... And like me, many users have also just mask udev to prevent any upgrading as a quick answer. And it's not really satisfying as i know this won't work long, even i saw you are making high efforts to push that delay farer.
That's why everyone is querying answers (ok, some just complain but we're users, complaining is our top skill)...
Back to top
View user's profile Send private message
Bones McCracker
Veteran
Veteran


Joined: 14 Mar 2006
Posts: 1564
Location: U.S.A.

PostPosted: Tue Feb 12, 2013 2:25 pm    Post subject: Reply with quote

Yamakuzure wrote:
Unification and simplification of related tasks is always better than separatism of the seemingly same into different parts. But some understand "the unix way" to mean "as complex as possible for the sake of it."

I think you may be missing the point. I haven't had enough coffee yet to care enough that to bother to explain it, though.

I'll just say this: architectural unification can be a good thing or a bad thing; it's certainly not "always" one or the other. It's entirely situation dependent. In fact, in the absence of compelling reasons for it, it's a useful rule of thumb to seek functional encapsulation and modularity. The level at which to expose that modularity is another question.

The architectural philosophy of UNIX is intended to benefit a more advanced class of users (system administrators, in essence) than that of MS Windows or Mac OS X -- users who need and benefit from the ability to combine a lower level of functional component -- people who do things such as write scripts and adapt to a wide variety of unforeseeable requirements.

This pattern of overly-amalgamated system architecture is characteristic of short-sighted focus on a limited set of use cases. I've seen it before in a variety of contexts. It is not admirable, clever, or elegant; it is sophomoric, lazy, and unprofessional. It comes out of being overly focused on facilitating desktop-like use cases and hiding system complexity from users. That's the kind of thinking that gave birth to "Microsoft Bob" and the albatross that the "Windows Registry" has become.

It's also an ineffective product introduction strategy I call the "ugly sister" method. If you want to date Charlene, you have to force your friend to take out her ugly sister Beula. The only people this makes happy are ugly Buela's parents, and even that is only in the short term.
_________________
True Liberals are individualists. Democrats, on the other hand, are authoritarian collectivists.
Back to top
View user's profile Send private message
ssuominen
Developer
Developer


Joined: 30 Sep 2005
Posts: 2108
Location: Finland

PostPosted: Tue Feb 12, 2013 3:06 pm    Post subject: Reply with quote

krinn wrote:

I would like to here from you, what you will do then ?


Maintain a minimal patchset that addresses the incompability issues (instead of doing complete fork) like gregkh suggested on the mailinglist earlier.
Back to top
View user's profile Send private message
ulenrich
Veteran
Veteran


Joined: 10 Oct 2010
Posts: 1136

PostPosted: Tue Feb 12, 2013 5:05 pm    Post subject: Reply with quote

@all, please consider how this paranoia in the Gentoo forum started: There was a maintainer at the edge of his capacity and workload, who stated he is not able to maintain any minor deviation from upstream systemd-udev. The followers of Gentoo~unstable then had experienced a little journey:

/lib/udev - /usr/lib/udev - /lib/udev
This provoked the feeling of misery and dependence and paranoia. Some of us didn't get the background. It is considered a bad act of courtesy a Gentoo developer speaking out all of it in public.

Udev just is a very little tool to transmit requests from user space to kernel space. We now have a cool udev maintainer stating (above) to easily manage Gentoo needs! And we have the eudev fork: A crew of people interested in an independend udev. I appreciate choice. I'm not that dumb fanboy of systemd.

@krinn, when assassing the commercial entity Redhat: They clearly are on the angel side in their playfield. What you talk about is the little community fedoraproject, assembled of people doing things they assume would uplift them to the commercial entity. Speaking about Gnome: The dependency on consolekit, which was a fedoraproject project anyway, was dismissed in favor of systemd. Nothing changed there in essence.
_________________
fun2gen2
Back to top
View user's profile Send private message
Yamakuzure
Veteran
Veteran


Joined: 21 Jun 2006
Posts: 1347
Location: Bardowick, Germany

PostPosted: Wed Feb 13, 2013 12:16 pm    Post subject: Reply with quote

BoneKracker wrote:
Yamakuzure wrote:
Unification and simplification of related tasks is always better than separatism of the seemingly same into different parts. But some understand "the unix way" to mean "as complex as possible for the sake of it."

I think you may be missing the point. I haven't had enough coffee yet to care enough that to bother to explain it, though.

I'll just say this: architectural unification can be a good thing or a bad thing; it's certainly not "always" one or the other. It's entirely situation dependent. In fact, in the absence of compelling reasons for it, it's a useful rule of thumb to seek functional encapsulation and modularity. The level at which to expose that modularity is another question.
Yes, I had thought this might go on like this without further explanation. But I posted (due to lack of time) against better judgment and apologize for that. ;)

What I mean with "Unification" is to put things together that belong together. These might as well be encapsulated and modularized tools. This is what the DEs come from. To have one collection of related (or not so related) tools ready in one go.

I know this is not the best example, but I can't think of a better (or more obvious) one right now. "kde-meta" pulls in everything (according to your USE flags) KDE related. But you are still able to use a web browser, mail program, text editor, file manager (and so on) of your choice.
Thus, KDE puts everything "they" think is necessary for a desktop environment together. And if you do not want to chose anything different, you just use what "they" provide.
This applies for gnome, xfce, enlightenment, razor-qt, lxde, and any other DE as well. The collection of tools might differ in number and functionality, but they all provide a fixed collection that you can change and/or substitute at will.

Now if systemd would just install a certain number of tools but let you chose to substitute one or another with something different (or do not install it at all) it wouldn't be this much different. Just for another topic, the system management, instead of desktop environment.

Or a different view: Just look at the gentoo handbook and put together all recommended tools like vixie-cron and syslog-ng and so on to be pulled in by a new package "sys-apps/gentood" - with a use flag for each tool. Anyone who just installs the handbook recommendations would need just one emerge and no second look.
But if you wanted, you could easily substitute syslog-ng with rsyslog for example.

As far as I understodd this, systemd does not offer this choice. And I have sympathy for those who feel highly affronted by that.
_________________
I *do* know that I easily aggravate people due to my condensed writing. Rule of thumb: If I wrote anything that can be understood in two different ways, and one way offends you, then I meant the other! ;)
Back to top
View user's profile Send private message
depontius
Advocate
Advocate


Joined: 05 May 2004
Posts: 2439

PostPosted: Wed Feb 13, 2013 5:14 pm    Post subject: Reply with quote

krinn wrote:
It we speak about having children, it's not off topic to speak about getting a wife :)

I'm sorry, I can't refuse this one... If we're speaking about having children, it's not off-topic to speak about urination, defecation, and potty-training, either. Now off the humor and back to something resembling technical/strategic discussion. I really wasn't casting aspersions, just indulging a little humor.
krinn wrote:
And i didn't test it, but this change looks strange : why would naming my network interface by bus place is better than mac ? There's more chance i will move my ethernet card from on slot to another than i would add another card with the same mac. High havock <> benefits ?

At my workplace we're currently working under a number of strategic directions. Multiple strategic directions doesn't have to be as insane as it sounds, because they apply at different levels of abstraction or aspects of operation, and don't necessarily conflict. (Necessarily is the bug, there. There is always a devil residing in the details)

Sometimes when a new strategy comes down from on high.
- If the bigwigs a layer or two down from that don't like it, they make the strategy's implementation awkward and annoying to the workers in the trenches. Eventually the hue and cry rises to management attention and the strategy gets modified.
- If the bigwigs a layer or two down from that do like it, but don't feel the power to absolutely mandate it, or know that it's going to cause a lot of pain, there is a different technique. They add onerous rules to working the old way, so you'll just give up and do it their way, even if it isn't as good, is horribly obnoxious, etc.

Where I work, both of the above are happening right now, with two (actually more) new strategies applying to two different aspects of the business.

In a way, systemd reminds me of this, where the cost of using udev in a non-strategic way is planned to become obnoxious and/or expensive. I'm probably mapping work politics into OSS, but it almost fits.
_________________
.sigs waste space and bandwidth
Back to top
View user's profile Send private message
hcaulfield57
Tux's lil' helper
Tux's lil' helper


Joined: 13 Mar 2012
Posts: 146

PostPosted: Thu Feb 14, 2013 4:29 am    Post subject: Reply with quote

krinn wrote:

No, i won't try or install systemd. I don't know if systemd is good or bad, and i don't care at all.

Same here. However, I'm quite sure that systemd works, and does what it's supposed to. But I don't care, OpenRC works and does what it's supposed to. The fact of the matter is, I use Linux as a hobby, and like to understand how my system works. Systemd adds an unnecessary layer of complexity and breaks with the existing UNIX philosophy of software design. There are some people who don't agree with the UNIX philosophy, and that's fine, but I'm convinced that minimalism, simplicity and elegance are what we should be looking for in software, rather than large opaque monoliths that don't work with existing tools on the system. When I regrettably have to use Windows, I'm always surprised with the unnecessary complexity it introduces just to get simple tasks done. Why adopt a new init system that breaks existing functionality? Everyone who wants to use systemd, more power to them, they have every right to do so, however a vision of Linux where one has to use systemd is contrary to the UNIX practice of modularity and FOSS software practices in general. As long as there remains a userbase that doesn't want to use systemd, I'm quite sure solutions will be around.
_________________
"To design the perfect anti-Unix, make all file formats binary and opaque, and require heavyweight tools to read and edit them." - The Art of Unix Programming
Back to top
View user's profile Send private message
Yamakuzure
Veteran
Veteran


Joined: 21 Jun 2006
Posts: 1347
Location: Bardowick, Germany

PostPosted: Thu Feb 14, 2013 8:58 am    Post subject: Reply with quote

Just another thought: I find it interesting that people advertising for systemd always state it "just works". For something this basic and fundamental it is the least I'd expect. And openrc+sysvinit does the same. It "just works". ;)
_________________
I *do* know that I easily aggravate people due to my condensed writing. Rule of thumb: If I wrote anything that can be understood in two different ways, and one way offends you, then I meant the other! ;)
Back to top
View user's profile Send private message
Display posts from previous:   
Reply to topic    Gentoo Forums Forum Index Gentoo Chat All times are GMT
Goto page Previous  1, 2, 3, 4  Next
Page 3 of 4

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum